Chief Technology Officer / India
The narrower claim still needed a consequential choice
- Company employees
- 180
- March 2026; twenty-four engineering colleagues through three leads
- Company revenue
- ₹76 crore
- FY 2025–26; ₹6.2 crore operating profit
- Industrial engineering experience
- 33 years
- Lighting development, applications and technology leadership
Decisions & consequences
01Holding the preproduction units after a longer run contradicted my claim
In 2019 I endorsed an intensity-stability statement for an industrial inspection-lighting assembly after a ten-minute check. The proposed application involved a longer operating run. When qualified colleagues reviewed forty-eight assemblies over ninety minutes under the defined conditions, twelve fell outside the project's agreed stability criterion. The short result had been real, but it did not establish the broader performance statement I had accepted.
Commercial colleagues proposed releasing the 240 preproduction units with an explanatory note and handling any concerns through applications support. I did not consider that adequate because the proposed claim still depended on a condition we had not supported. We held the units and funded a ten-week, ₹46 lakh programme of design review, correction and retesting. Quality head Vikas Pelnqeri retained acceptance authority rather than treating my sponsorship of the programme as a release decision.
The repeat review used the same defined ninety-minute criterion and the same forty-eight-unit group after the correction. Forty-five met it; three remained held for further work. I reported those three separately and did not describe the result as all units qualified. The programme did not establish lifetime, behaviour in every ambient condition or the performance of a customer's complete inspection process. The account concerns preproduction qualification; it reports neither a customer recall nor a personal injury.
The delay affected the introduction plan and required a revised conversation with prospective customers. I explained that my earlier approval had allowed a convenient test interval to carry a claim beyond its demonstrated scope. The team should not have had to hide the longer-run failures to defend that approval. We changed the development review so that the period and operating conditions accompanied the proposed statement before commercial wording was agreed.
The outcome was a narrower, better-supported account of the assembly and an explicit hold on the remaining cases. I learned that acknowledging a technical limit can preserve the usefulness of a result. The alternative is to let a short successful demonstration become a promise that engineers and quality colleagues are later expected to justify under conditions they never claimed to have examined.
Qualifying the camera combinations instead of calling the light universally steady
A 2023 evaluation covered thirty-two camera and lighting configurations for a proposed inspection offer. The light appeared steady when observed directly, yet nine configurations showed banding against the project's defined image criterion. The development discussion initially treated visual steadiness as reassuring evidence. Applications colleagues showed that the camera's exposure and readout interacted with the lighting control in ways the observer's impression did not describe.
I supported an eight-week, ₹58 lakh programme to examine the relevant timing and control arrangements. The work involved qualified applications, electronics and quality colleagues, with the configuration and settings recorded for each case. I did not issue a universal operating prescription or ask commercial teams to describe a control change as solving every camera interaction. The purpose was to establish what we could support in the defined offer.
After the programme, twenty-seven of the thirty-two configurations met the agreed image criterion. Five remained outside the qualified scope. The commercial team wanted a broader statement because customers preferred a simple compatibility message. I supported publishing the examined combinations and their limits instead. A customer using another configuration would need an appropriate application review; the result could not be transferred merely because its camera looked similar in a catalogue.
The narrower offer required more explanation and deferred opportunities that depended on the five unresolved configurations. We did not count those opportunities as orders or claim a measured increase in inspection accuracy across customer factories. Product acceptance and the customer's inspection decisions remained with their authorised specialists. The lighting evidence was one part of that wider assessment.
The case strengthened my view that the unit of technical judgement is sometimes the combination, not the component. A light, camera and control setting can each function as designed while their interaction fails to produce the intended image. My responsibility was to support the work needed to understand that interaction and keep its boundary visible when the result moved from an engineering review into a commercial claim.
Choosing a qualified supplier controller while accepting its higher unit cost
In 2025 the engineering team proposed developing a controller internally for a new inspection-lighting range. The plan required ₹2.8 crore and twelve months. A supplier-module route required ₹1.1 crore of qualification and integration work over four months, but its quote carried an additional ₹1,600 per unit compared with the internal recurring-cost estimate. At a forecast 1,200 units a year, that difference implied ₹19.2 lakh annual variable cost. The forecast was not a committed order book.
Some colleagues preferred the internal design because it would keep more engineering decisions within our team. That was a real advantage to consider, but it did not remove the development effort or the support obligation after launch. I asked for a comparison of the required functions, qualification scope, supplier support, permitted use and the consequences of future component changes. Commercial and professional colleagues reviewed the arrangements relevant to their authority.
I recommended the supplier route for the defined range, accepting the higher quoted unit cost in exchange for the narrower initial programme and earlier qualification opportunity. The CEO approved ₹1.1 crore for that scope. The four-month programme qualified four of six lighting variants; two remained outside the offer pending further work. We did not present the approval as proof that all six were ready, or that the supplier route would always be cheaper over the product's life.
The comparison kept development spending, elapsed time and forecast recurring cost separate. We did not subtract one proposal from the other and call the difference an achieved saving without considering scope. Nor did the ₹19.2 lakh estimate establish a complete investment-payback period. Actual volumes, support needs and later changes would affect the economic result, and supplier dependence remained part of the ongoing account.
The decision preserved engineering attention for the lighting application rather than making ownership of every controller component the objective. I learned that a CTO can support an external technology choice without surrendering technical responsibility. The responsibility shifts toward qualification, interfaces, supplier change and honest communication of the resulting limits. Internal invention should earn its place through the problem it solves, not through the pride of keeping every design decision inside the company.